Fritz,

> Tom,
> I am not sure if understand the trial system of FDT250x:  
> The kernel will be online for one month till the next version e.g. FDT2507. 
> So if someone reports a bigger problem with the kernel it will disappear 
> within one month.  
> I know bugs that are 20 and more years old.  
> As nobody worries about the bug report list, this is the fastest way to find 
> problems.  
> And it is very easy to test this FDT on a 500 GB real HD with a 2 TB HD in a 
> virtual machine. It is installed within 10 minutes and, if necessary, deleted 
> in two minutes.  
swapping kernel.sys between old and new version takes ~5 seconds.

> So nobody can say that this "a lot of work" to test. It needs less time than 
> writing a mail.  
> I am aware than virtual machines do not cover everything, but I would say 
> more than 80%. In doubt it is always good to test with two different vms and 
> a real machine.  
>    
> Till now I saw a lot of benefits of the new kernel (tested today):  
>    
> here some of them:  
> - support for creating real 2 TB images (fdisk already did it, but now format 
> also supports it),  
> - support for creating a small FAT-16 HD behind a big FAT32 one (e.g. a third 
> HD) - format gave an error till now, now I tested 2 x 2 TB and two smaller 
> HDs.  
> - your max 4 GB support for the copy command now works,  
> - fdisk works on a 4 TB HD, but of course, supports max. 2 TB,  
> - if exist Z:\NUL echo drive Z exists now works again, etc.  
I do not doubt that the new kernel has a lot of advantages.

I just wonder why it hasn't been published ages ago.
   
> So why not giving it a trial for one month? 

> New iso files make it easier to create a new machine than adding single 
> kernels or other files via diskette / vbados or imdisk. 
If you need a new release to get a changed kernel.sys to your machine, you are 
doing something wrong.

All that said, just go ahead. Good luck.

Tom


>  
>    
> Of course there is nothing to say against an update of this kernel if 
> something goes wrong or is still missing.  
>    
> Willi  
>    
>    
>    
>   
>   
> Sent: Tuesday, May 27, 2025 at 4:17 PM
> From: "tom ehlert via Freedos-devel" <freedos-devel@lists.sourceforge.net>
> To: "Technical discussion and questions for FreeDOS developers." 
> <freedos-devel@lists.sourceforge.net>
> Cc: "tom ehlert" <t...@drivesnapshot.de>
> Subject: Re: [Freedos-devel]  Re: Kernel package question  
> Hallo Herr Bernd Böckmann via Freedos-devel,

>  am Dienstag, 27. Mai 2025 um 11:52 schrieben Sie:


 >>> Am 27.05.2025 um 09:02 schrieb tom ehlert via Freedos-devel 
 >>> <freedos-devel@lists.sourceforge.net>:
 >>>
 >>> A new kernel.sys can be - and should be - tested independently of all the 
 >>> surrounding stuff,
 >>> so you have potential issues automatically isolated to the kernel, and 
 >>> nothing else.

 >> It is not that the new kernel is completely untested. For instance, the 
 >> links to recent kernel binaries were provided on this list. Everyone with 
 >> interest had the opportunity to test it. Sadly, not many did so. But a few, 
 >> like Willi, invested a significant time testing it. So including it in the 
 >> test release is the next logical step to widen the group of testers, which 
 >> nevertheless I'm afraid won't be many...

>  I don't have any secret information.
>  I'm just worried that there might be a reason PerditionC hasn't released 
> *any* kernels over the last few years, while there has been so much 
> added/fixed.
>  And I have asked *often*. Nothing. I'm not sure taking the kernel "as is" 
> from GitHub anyway is such a great idea.

>  Kernel 2037 2.0?

>  Tom



>  _______________________________________________
>  Freedos-devel mailing list
>  Freedos-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/freedos-devel





Mit freundlichen Grüßen / with kind regards
Tom Ehlert
+49-15151898538



_______________________________________________
Freedos-devel mailing list
Freedos-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/freedos-devel

Reply via email to